Orientador: Paulo de Oliveira Camargo / Resumo: Assim como no posicionamento por ponto, os resultados obtidos no posicionamento relativo a partir de receptores de uma freqüência são afetados pelo efeito sistemático da ionosfera, que atualmente é uma das maiores fontes de erro no posicionamento com GPS. O problema principal no posicionamento relativo é devido ao gradiente espacial do conteúdo total de elétrons (TEC) presente na ionosfera, enquanto que no posicionamento por ponto o principal problema é a magnitude do TEC. Assim, no posicionamento relativo fica mais complicado de se analisar o efeito da ionosfera nos resultados obtidos, principalmente na região equatorial, onde várias irregularidades e anomalias ocorrem. Este projeto analisa o efeito da ionosfera no posicionamento relativo estático na região equatorial, em diversas condições ionosféricas, a partir de uma abordagem teórica e prática. Foram processados dados somente da portadora L1 com a dupla diferença da fase e do código. A análise dos resultados foi realizada a partir das discrepâncias das coordenadas consideradas verdadeiras com as obtidas no processamento das linhas de base. Três experimentos foram realizados, no qual foi possível analisar a influência da ionosfera no posicionamento relativo a partir de vários aspectos. De uma forma geral, verificou-se na prática a complexidade de definir se determinada linha de base é curta ou média, devido ao gradiente espacial do TEC. Também foram verificados efeitos da cintilação ionosférica e de uma tempestade geomagnética nos resultados obtidos. / Abstract: The obtained results in the relative positioning, as well as in the point positioning, are affected by the ionosphere systematic effects, which is one of major error sources in the GPS positioning. The main problem in the relative positioning is due to the Total Electron Content (TEC) variation of the ionosphere, while in the point positioning the main problem is the magnitude of the TEC. Thus, the effects of the ionosphere in the relative positioning are more difficult to analyze, mainly in the equatorial region, where several irregularities and anomalies occur. This research analyzes the ionosphere effect in the relative positioning in the equatorial region considering several ionosphere conditions, both from theoretical and practical standpoints. Only L1 carrier data have been processed using L1 phase and C/A-code doubledifferences. The analyze of the obtained results have been carried out from the discrepancies between the "true" coordinates and corresponding ones obtained in the processing of the baselines. Three experiments have been accomplished, in which it was possible to analyze the ionosphere influence in the relative positioning considering several aspects. In general, due to TEC variations, it has been verified, in practice, the complexity of defining whether a given baseline length is short or medium. It has been also verified in the results the effects of the ionospheric scintillation and of a geomagnetic storm. / Mestre

HMK Geodesi, GPS publicerades år 1996 och är fortfarande det regelverk som idag tillämpas vid statisk mätning med GNSS. Studien genomförs för att bidra med rekommendationer till nya HMK-Stommätning för statisk GNSS-mätning.   Studiens syfte var att undersöka baslinje- och sessionslängdens påverkan på lägesosäkerheten vid mätning av korta baslinjer. Eftersom studien skulle efterlikna mätning under praktiska förhållanden undersöktes baslinjelängder 0,7–100 km och sessionstider 20 min–6 h. Syftet var också att undersöka om lägesosäkerheten påverkas olika vid beräkning med de olika frekvenserna, L1, L1+L2 eller L3.   Två delstudier genomfördes i två olika geografiska områden. I Gävleområdet användes data från egna mätningar i kombination med data från en SWEPOS-station för beräkning av spridningen i position för korta baslinjer. I Göteborgsområdet har data erhållits från och beräknats mellan 14 SWEPOS-stationer. En felvektor har beräknats mellan stationens beräknade position i studien och en given position beräknad av SWEPOS.   I Gävleområdet var skillnaden i standardosäkerhet för koordinaterna mellan olika sessionstider mindre än 3 och 7 mm i plan respektive i höjd. För baslinjer upp till 5 km är spridningen i höjd i hälften av fallen mindre än i plan för frekvenserna L1, L1+L2 och L3. Vid längre sessioner gav frekvenserna likvärdiga resultat. I Göteborgsområdet ökade höjddifferensen i samband med baslinjelängden. Den tredimensionella avvikelsen för baslinjerna var 1–71 mm.   Lägesosäkerheten i plan påverkas inte nämnvärt av ökad baslinjelängd. Lägesosäkerheten i höjd blir större när baslinjelängden ökar. Till skillnad från tidigare studier ses ingen tydlig förbättring när sessionstiden ökar. / HMK Geodesy, GPS was published in 1996 and is still the regulations that currently apply to static measurement with GNSS. The study is conducted to provide recommendations for new HMK Control networks for static GNSS surveying.   The purpose of the study was to investigate how the baseline and session length influence the position uncertainty when measuring short baselines. A second purpose was to investigate whether there will be any difference in the position uncertainty when using different frequencies: L1, L1 + L2 or L3. The study investigates baseline lengths 0,7–100 km and session lengths 20 min–6 h. The study was designed to mimic measurements during practical conditions and investigate the position uncertainties that can be expected to be achieved by users.   Two sub-studies were conducted in two different geographical areas. In the Gävle area self-produced data was used combined with data from a SWEPOS station for calculating the spread in the position for short baselines. In the Gothenburg area the data was obtained from observations of known SWEPOS reference stations. An error vector was calculated between the station's estimated position of the study and a given position calculated by SWEPOS.   In the Gävle area the differences in standard uncertainty between different session lengths for the coordinates were less than 3 mm and 7 mm in plane and height. For baselines up to 5 km the spread in height in half of the cases was lower than in plane for the frequencies L1, L1+L2 and L3. During longer sessions the results for the different frequencies are equal. In the Gothenburg area the height differences increased when associated/combined/correlated with baseline length. The three-dimensional deviations of the base lines were 1–71 mm.   The position uncertainty in plane is not affected significantly by increased baseline length. Location Uncertainty in height becomes larger when baseline length increases. Unlike previous studies our study showed no clear improvement in position uncertainty when session length increases.

Made available in DSpace on 2014-06-11T19:23:30Z (GMT). No. of bitstreams: 0 Previous issue date: 2005-09Bitstream added on 2014-06-13T20:50:26Z : No. of bitstreams: 1 dalpoz_wr_me_prud.pdf: 2438411 bytes, checksum: 8c56dfb5193a59fce51aa2a4633b24a1 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Assim como no posicionamento por ponto, os resultados obtidos no posicionamento relativo a partir de receptores de uma freqüência são afetados pelo efeito sistemático da ionosfera, que atualmente é uma das maiores fontes de erro no posicionamento com GPS. O problema principal no posicionamento relativo é devido ao gradiente espacial do conteúdo total de elétrons (TEC) presente na ionosfera, enquanto que no posicionamento por ponto o principal problema é a magnitude do TEC. Assim, no posicionamento relativo fica mais complicado de se analisar o efeito da ionosfera nos resultados obtidos, principalmente na região equatorial, onde várias irregularidades e anomalias ocorrem. Este projeto analisa o efeito da ionosfera no posicionamento relativo estático na região equatorial, em diversas condições ionosféricas, a partir de uma abordagem teórica e prática. Foram processados dados somente da portadora L1 com a dupla diferença da fase e do código. A análise dos resultados foi realizada a partir das discrepâncias das coordenadas consideradas verdadeiras com as obtidas no processamento das linhas de base. Três experimentos foram realizados, no qual foi possível analisar a influência da ionosfera no posicionamento relativo a partir de vários aspectos. De uma forma geral, verificou-se na prática a complexidade de definir se determinada linha de base é curta ou média, devido ao gradiente espacial do TEC. Também foram verificados efeitos da cintilação ionosférica e de uma tempestade geomagnética nos resultados obtidos. / The obtained results in the relative positioning, as well as in the point positioning, are affected by the ionosphere systematic effects, which is one of major error sources in the GPS positioning. The main problem in the relative positioning is due to the Total Electron Content (TEC) variation of the ionosphere, while in the point positioning the main problem is the magnitude of the TEC. Thus, the effects of the ionosphere in the relative positioning are more difficult to analyze, mainly in the equatorial region, where several irregularities and anomalies occur. This research analyzes the ionosphere effect in the relative positioning in the equatorial region considering several ionosphere conditions, both from theoretical and practical standpoints. Only L1 carrier data have been processed using L1 phase and C/A-code doubledifferences. The analyze of the obtained results have been carried out from the discrepancies between the true coordinates and corresponding ones obtained in the processing of the baselines. Three experiments have been accomplished, in which it was possible to analyze the ionosphere influence in the relative positioning considering several aspects. In general, due to TEC variations, it has been verified, in practice, the complexity of defining whether a given baseline length is short or medium. It has been also verified in the results the effects of the ionospheric scintillation and of a geomagnetic storm.

GNSS är idag den vanligaste tekniken för positionering och för att ansluta punkter till det nationella referenssystemet SWEREF 99. Systemet realiseras av ett nätverk av permanenta referensstationer, SWEPOS. Nätet består av ett mindre antal stabilt förankrade klass A-stationer och ett större antal, men mindre stabila klass B-stationer. Lantmäteriet erbjuder idag tre tjänster för efterberäkning av GNSS-data som använder SWEPOS-nätet: 1) SWEPOS Beräkningstjänst, som beräknar positioner genom anslutning till klass A-stationer med långa baslinjer som följd, 2) nedladdning av RINEX-data från valfria SWEPOS-stationer, vilket möjliggör medellånga baslinjer och 3) skapande av virtuella referensstationer (VRS), där baslinjerna är mycket korta. Syftet med detta examensarbete är att, genom att använda dessa tjänster, jämföra olika metoder för anslutning av statisk efterberäknad GNSS-mätning till SWEREF 99 och un-dersöka hur mätosäkerheten beror av sessionstid och baslinjelängd. Dessutom undersöks påverkan av mätmiljö och om det finns någon systematisk skillnad mellan de olika metoderna. Fyra s.k. SWEREF-punkter användes som kontrollpunkter. Punkterna mättes i tre 8 h-sessioner. Insamlat mätdata delades in i fönster och skickades till SWEPOS Beräkningstjänst, beräknades i en kommersiell programvara mot klass A-, närmaste stationer och en VRS. För att jämföra de olika metoderna beräknades RMS och standardosäkerheter. Resultaten visar att vid kortare sessionstider har SWEPOS Beräkningstjänst ett högre RMS än övriga metoder, men vid lätt och normal mätmiljö sjunker RMS till samma nivå som övriga metoder redan efter 1 h sessionstid, vilket var 1 cm eller lägre i plan och 2 cm eller lägre i höjd. Vid svår mätmiljö fortsätter RMS att sjunka ända upp till 4 h sessionstid och ligger då något högre än vid lätt och normal mätmiljö. För övriga metoder sjunker RMS inte lika tydligt när sessionstiden ökar. Det framgår att mät-miljön påverkar mätresultatet. Både standardosäkerheter och RMS är högre vid svår mätmiljö jämfört med lätt och normal mätmiljö. Signifikanta skillnader mellan de olika metodernas medelavvikelser kunde påvisas, vilket innebär att det finns systema-tiska avvikelser mellan metoderna, som kan bero på olika troposfärsmodeller och att SWEPOS beräkningstjänst använder referenssystemet ITRF för att sedan göra en transformation till SWEREF 99. / GNSS is at present the most frequently used method for positioning, as well as connecting new points to the Swedish national reference frame, SWEREF 99. The reference frame is realized by a network of permanent reference stations, SWEPOS. The network consists of a small number of rigidly mounted class-A stations and a larger number of less stable class-B stations. Lantmäteriet, the mapping, cadastral and land registration authority of Sweden, cur-rently offers three services for post processing of GNSS data that utilizes the SWEPOS network: 1) SWEPOS Post Processing Service, which computes coordinates by connecting baselines to class-A stations leading to long baselines. 2) Downloading of RINEX data from any SWEPOS stations leading to medium baseline lengths, and 3) creation of virtual reference stations (VRS), leading to very short baselines. The aim of this thesis is, by using these services, to compare different methods for connecting new points to SWEREF 99, using post processing of static GNSS measurements and to investigate the impact of session duration and baseline length on the uncertainty of the measurements. The impact of different measurement environments and systematic effects between the methods are also investigated. Four SWEREF-points were used as test points. The points were measured in three 8 h sessions. The data were divided into windows and sent to SWEPOS post processing service, as well as processed with a commercial software, where baselines were pro-cessed against class-A, class-B stations and a VRS. To compare the different methods RMS and standard uncertainties were calculated. The results indicate that shorter session duration yields higher RMS for SWEPOS post processing service when compared with the other methods, but in easy measurement environments RMS decreases to the same level as the other methods after 1 h session duration, which is 1 cm or less horizontally and 2 cm or less vertically. However, in complex measurement environments RMS continues to decrease up to 4 h session duration and is higher than in easy environments. For the other methods the decrease of RMS is not as pronounced. It is clear that the measurement environment impacts the results. Both standard uncertainty and RMS are higher in complex environments compared with easy and moderate environments. Significant differences of the mean deviation for each method could be detected, which indicates systematic effects between the methods, that could depend on different troposphere models and that SWEPOS Post Processing Service uses the reference system ITRF and then make a transformation to SWEREF 99.